Abstract: Detecting and monitoring plant diseases are a crucial task in agriculture. The agricultural sector plays a vital role in the Indian economy, and early detection of crop diseases can improve both the quantity and quality of crops while reducing the time required for disease detection. This system for pepper plant leaf disease detection begins with data collection, which can be done through either surveying local agricultural land or collecting data from agricultural colleges. The collected images are then pre-processed to suppress unwanted distortions and enhance image features like contrast and size. Feature extraction is performed using the discrete wavelet transform, particularly the Haar wavelet compression, which samples the wavelet at discrete intervals. An artificial neural network classifier then classifies the image as healthy or unhealthy by utilizing the extracted image features. The classifier employs the back-propagation algorithm to achieve higher performance and accuracy.
Keywords: Artificial Neural Network (ANN), DWT2 (2-D Discrete Wavelet Transform), Image Processing, Pepper Plant
|
DOI:
10.17148/IJARCCE.2023.124138
